Quote by ghwellsjr First you say your question is not related to the string scenario and then you proceed to exactly describe the string scenario, except that it is replaced by a rigid rod, and then you agree that obviously there is going to be stress. So I'm not sure what you are looking for. Nevertheless, even though this issue has been dealt with countless times in this and other threads, I will say succinctly that if you accelerate the two ships identically then they will maintain the same distance apart in their initial rest frame. But the connecting rod between them will be subject to length contraction in the same initial rest frame. Therefore, if it is rigid, it will break, if it is not rigid, it will stretch.
Yes this topic has come up before. A long time ago I had a prolonged discussion just like this, with someone who maintained the same basic view that you hold. Unfortunately that discussion got derailed midstream into a side argument about whether conservation of momentum could be applied absolutely in the real world. In any case the main question was never resolved so I welcome this opportunity to explore it.

Originally Posted by ghwellsjr

 Yes, if we move a rod by accelerating it at one point then we won't squeeze or pull on it but if we accelerate one end of rod separately from accelerating the other end of the rod, we can end up squeezing it or pulling it apart. Isn't that obvious?
I am sorry but it is not obvious to me. You say that acceleration from one end does not cause a problem or disruption. (SO in this case that would mean only firing up the lead ship). But application at two points would result in disruption.
If the force is applied only at the front, that creates the maximum expansive stress possible without applying reverse thrust to the rear. SO adding a forward thrust at the rear actually reduces the overall expansive stress, so I am confused as to why you think this would lead to expansive disruption where the single thrust would not.

The basic premise of Born rigid acceleration; that stressless acceleration would necessitate a scaled and distributed acceleration scheme is certainly reasonable.
But aside from the fact that it is unrealistic in application , until we develop some totally new science that negates inertia and momentum,(gravity drive or???) stress is an inevitable consequence of acceleration, and stress, per se, is not a big problem. We live every day under a constant stress of 1 g. The relevant concern is if that stress is constant or dynamically increasing. So why do you think it would be increasing to the point of disruption?
You also did not provide any basis for your assumption that equal thrust at the front and rear would necessarily result in equal coordinate acceleration at those points and a constant separation in the launch frame.

 Quote by yuiop Thanks for the agreement. I cannot think of a clearer demonstration of the physical nature of length contraction in SR. Surely that woud depend on the power of the rockets and the tensile strength/ elasticity of the connecting tether. For example if the connector was a bungee cord then a pair of powerful rockets would have no trouble maintaining constant separation in the launch frame and the elasticity of the bungee can take a lot of stress without breaking, but of course with indefinite acceleration, it would have to eventually snap.
i was talking within the context of this discussion; assuming a realistically strong rigid connecting structure and reasonable acceleration. Certainly the actual stresses and results would be affected by the magnitude of acceleration. I would imagine that given a sufficiently long structure and high enough magnitude of acceleration, that serious deformation ,even to the point of disruption could occur in the front section before the momentum reached the middle and motion began at that point. But what if it was a dynamically increasing acceleration, starting from 0 and slowly increasing to the desired final magnitude??

austinO post 191
 I myself find the, physical contraction as a consequence of EM and atomic light speed interactions hypothesis very convincing. But as you have shown here it is somewhat problematic in application to specific scenarios.
This seems to be the same problem I was working on recently. Maybe it will help you.
U is the universal rest frame. A and B space ships pass U at t=0, moving at v = .6c.
Both experience equal length contraction to .8L in the x direction. If length contraction
is a result of em deformation in response to acceleration, then length expansion should
be the response to deceleration. If the A ship returns to U and stops, it should recover
its original length.
According to SR, if A moves away from B, B should measure a length contraction of A.
At first it seems A would have to expand and contract simultaneously to satisfy both
requirements, but not so. The confusion occurs because there are two different length
contractions, the first due to absolute motion relative to light speed, the second due
to perception. Since U is the absolute rest frame, the A & B contraction is the result of
em phenomena. After deceleration of A to v=0, it expands to 1L. Now consider B as
passing A at rest in the U frame. Time dilation for B is .8t, thus B arrives early at
locations on the Ux axis. Since everything in the B frame slows B trusts his clock and
interprets the effects as length contraction of the U frame, thus A is contracted to
.8L, and both requirements are met.